Optimization Method Based On The Composition Predicted Sintering Ingredients Experts

The sinter mine is one of the main materials of the blast furnace. The method of sinter is widely used because of it's comparably low cost. It is very important to the iron and steel industry. In the former process of sinter, several kinds of steel mines, solvents and fuels were mixed together with a certain ratio. The veracity of the mixing ratio affects the quality of the sinter mine. It will also have effects on the process of smelt in the blast furnace.The optimized control target of the mixing is to ensure the veracity and stability of the sinter mine. To solve problems of the long time interval of detection and lack of theory guidance of adjustment and the component of the mine fluctuate too much, this paper proposes an expert control strategy based on ingredient prediction. The process of mixing is divided into neutralizing mixing, blending mixing and form of the sinter mine. Three predicting models were established according to the strategy. Mathematic model and neural networks were adopted. On the aspect of the adjustment, expert rules were established according to the mechanism analysis and knowledge of expert. The rules can be used to adjust the mixing ratio of the mines .In order to decrease the cost of the sinter mine, linear programming were used. The optimized control of sintering can be implemented by the combination of mathematical model, neural networks and expert rules.To verify the applicable of the strategy, this paper uses the real data of the 105m2 sinter machine of shaoguan iron and steel plant. The result of the simulation shows that the prediction models can describe the process of sinter properly and the expert rules does works. The strategy proposed by the paper can be put into application with no doubt.